The problem: when order management doesn't scale#
Managing webshop orders seems simple when traffic is low. But as soon as dozens of orders come in daily, manual work quickly spirals out of control. You need to figure out who paid and who didn't. Which order goes to whom for packing. Who needs to be notified. Which video product needs to be sent out. Which customer gets a reminder.
At the Mindwell webshop, this was exactly the situation. The team was spending hours every day manually managing orders. Shoprenter, Airtable, email, phone calls. Everything scattered, everything done by hand. If a team member was sick, nobody knew exactly where a given order stood.
What made things more complicated was that Mindwell doesn't just sell physical products: they also offer video content in their webshop, which needs to be delivered to the buyer via a unique link after purchase. This is a completely different process from physical delivery, and both had to be managed in parallel.
The approach: many small workflows, not one big system#
We didn't build a monumental, all-in-one system. Instead, we created 11 smaller, well-defined n8n workflows that communicate with each other. This approach has serious advantages:
- If one workflow fails, the others keep running
- Each workflow can be tested and fixed independently
- New features can be added one at a time
- The team can see exactly what each automation does
For the Mindwell project, we chose n8n because:
- It runs on our own server, so there's no vendor lock-in and the data stays with the company
- No execution limits: Make.com and Zapier charge per execution, n8n doesn't
- More flexible logic: any branching, loops, and error handling can be implemented in n8n
- Native webhook handling: it receives Shoprenter webhooks directly
System architecture#
The 11 workflows cover three main areas: order routing, video product sales, and data sync. Let's look at how each one works in detail.
1. Order routing: the smart dispatcher#
This is the heart of the system. When an order comes in through Shoprenter, it notifies n8n via webhook. The first workflow examines the order data and automatically decides what should happen next.
Based on payment method:
- Bank transfer: the order goes to "Awaiting payment" status, the system monitors the bank account, and automatically proceeds when the transfer arrives
- Cash on delivery: the order immediately goes to "Ready for packing" status, since payment happens at delivery
- Online payment (card): proceeds immediately after Shoprenter confirms the payment
Based on shipping method:
- Courier service: the system prepares the shipping data
- Personal pickup: sends a notification email with pickup point details
- Parcel point: records the selected parcel point details
Based on product type:
- Physical product: goes to the appropriate warehouse packing task list
- Digital product (video): triggers the video delivery workflow
- Mixed order: both processes start in parallel
Based on these factors, the order automatically moves to the right status. No human decision, no errors. The team only sees the task list that's relevant to them.
2. Video products: fully automated digital delivery#
Mindwell also sells video content through their webshop. This is a special area where the value of automation is particularly visible, because the entire process runs without human intervention.
Video preparation (one-time)
The Mindwell team uploads new video content to Google Drive. This is the only manual step in the entire process. The system checks daily whether there's a new video in the designated folder.
Automatic processing at purchase
When someone buys a video product, the system automatically:
- Downloads the video from Google Drive
- Uploads it to Cloudflare Stream (which provides fast, global video streaming)
- Generates a unique, time-limited playback link for the buyer
Delivery via email
The buyer receives an automatic email with the playback link and a customized message. The email includes:
- The video title and description
- The unique playback link
- The expiration date (30 days)
- Technical support info (which browsers and devices are supported)
Reminders and expiration
The system automatically:
- Sends a reminder email 7 days before expiration
- Sends a final reminder 2 days before expiration
- On expiration, deletes the video from Cloudflare and deactivates the link
Error handling. If something goes wrong at any step (for example, the Cloudflare upload times out), the team receives a Telegram message with error details, and the workflow retries. The system resolves most errors on its own; human intervention is rarely needed.
3. Airtable sync: the central database#
Airtable is Mindwell's "control tower." Every order, customer, and item automatically appears here in real time. But this isn't simple data copying: the sync works intelligently.
At the order level:
- Every Shoprenter order appears in Airtable with complete data
- Status changes sync both ways (if something is modified in Airtable, Shoprenter updates too)
- The system calculates turnaround times (from order to delivery)
At the customer level:
- Every buyer is automatically added to the customer registry
- The system tracks purchase history
- For returning customers, previous orders are also visible
At the item level:
- Every ordered product appears on a separate row
- Inventory updates automatically
- If a product's stock drops below a critical level, the team gets a notification
Airtable is a great choice for a system like this because:
- The team can filter, sort, and create reports without programming knowledge
- Views can be configured as needed (packing view, finance view, customer service view)
- Additional logic can be built in using automations and formulas
- It's accessible from mobile, which is especially useful for warehouse work
Implementation steps#
We didn't build the entire system at once. The rollout happened in three phases, each delivering value on its own.
Order routing (Week 1)
First, we automated the order routing: automatic categorization and status management for incoming orders. This alone replaced 1-2 hours of manual work per day, because the team no longer had to review orders one by one.
Video sales (Weeks 2-3)
Once routing was running stably, we moved on to automating video delivery. This was the most complex part because it involved connecting multiple external services (Google Drive, Cloudflare Stream, email sending). Testing took the most time here.
Airtable sync and reporting (Week 4)
Finally, we built the Airtable sync, which was the most visible change for the team. From this point on, they could see everything in one place: orders, customers, inventory, tasks.
Results in numbers#
| Metric | Before | After |
|---|---|---|
| Manual work per order | 10-15 min | 0 min |
| Daily admin time | 3-4 hours | less than 30 min |
| Order processing time | 1-2 hours | real-time |
| Video delivery | manual email, 1-24 hours | automatic, within minutes |
| Orders with wrong status | 3-5 per week | virtually zero |
| Number of workflows | 0 | 11 |
| Automated steps per order | 0 | 30+ |
The Mindwell team now spends their freed-up time on what actually matters: content creation, customer relationships, and development. Administration is no longer a bottleneck.
Tech stack#
| Tool | Role |
|---|---|
| n8n | The engine behind all automations: webhook handling, logic, error management |
| Shoprenter | Webshop platform, source of orders and products |
| Airtable | Central database, task management, reporting |
| Cloudflare Stream | Video storage, streaming, time-limited links |
| AWS SES | Email notifications (order confirmation, video links, reminders) |
| Google Drive | Source folder for video content |
| Telegram | Error notifications to the team |
| Docker | Isolated execution of the video processing service |
Lessons learned#
You don't have to automate everything at once. We built the Mindwell system step by step. Each phase delivered value on its own, and the team had time to get used to the new workflow.
Error handling is not optional. Any automation is only as good as its ability to handle errors. Every workflow has retry logic, and if something still breaks, the team gets notified immediately.
Involving people in the design is key. We designed the Airtable views together with the packing team, because they knew what data they needed for their daily work.
Related articles#
If you're interested in automation and want to implement it in your own company, these articles may be useful:
- 5 processes any SMB can automate - you don't have to start with 11 workflows, even one can make a big difference
- No-code vs. custom development: which should you choose? - in the Mindwell project, we used a combination of both